admin-assistant

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect prompt injection vulnerability surface identified due to the ingestion of untrusted external data.
  • Ingestion points: External data enters the context through Google Cloud Storage (GCS) imports, third-party SaaS connectors (Jira, Salesforce, Confluence), and target site URL patterns for web crawling (SKILL.md).
  • Boundary markers: Absent. The instructions do not define delimiters or specific constraints to prevent the agent from executing instructions potentially embedded within the ingested data.
  • Capability inventory: The skill provides instructions for tools that can create, delete, and purge data stores, apps, and documents, as well as assign or unassign user license seats (SKILL.md).
  • Sanitization: Absent. No specific filtering, validation, or sanitization logic is mentioned for content retrieved from external websites or third-party platforms.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 12:21 PM
Security Audit — agent-trust-hub — admin-assistant